CN113777610B - Low-power-consumption underwater acoustic target detection system suitable for small platform - Google Patents

Low-power-consumption underwater acoustic target detection system suitable for small platform Download PDF

Info

Publication number
CN113777610B
CN113777610B CN202111348394.6A CN202111348394A CN113777610B CN 113777610 B CN113777610 B CN 113777610B CN 202111348394 A CN202111348394 A CN 202111348394A CN 113777610 B CN113777610 B CN 113777610B
Authority
CN
China
Prior art keywords
module
target
detection
frequency
cluster
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202111348394.6A
Other languages
Chinese (zh)
Other versions
CN113777610A (en
Inventor
李建龙
蒋丞
颜曦
杨志国
徐文
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Rayfi Technology Co ltd
Zhejiang University ZJU
Original Assignee
Hangzhou Rayfi Technology Co ltd
Zhejiang University ZJU
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Rayfi Technology Co ltd, Zhejiang University ZJU filed Critical Hangzhou Rayfi Technology Co ltd
Priority to CN202111348394.6A priority Critical patent/CN113777610B/en
Publication of CN113777610A publication Critical patent/CN113777610A/en
Application granted granted Critical
Publication of CN113777610B publication Critical patent/CN113777610B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
    • G01S15/02—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems using reflection of acoustic waves
    • G01S15/04—Systems determining presence of a target
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S1/00—Beacons or beacon systems transmitting signals having a characteristic or characteristics capable of being detected by non-directional receivers and defining directions, positions, or position lines fixed relatively to the beacon transmitters; Receivers co-operating therewith
    • G01S1/72—Beacons or beacon systems transmitting signals having a characteristic or characteristics capable of being detected by non-directional receivers and defining directions, positions, or position lines fixed relatively to the beacon transmitters; Receivers co-operating therewith using ultrasonic, sonic or infrasonic waves
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S15/00—Systems using the reflection or reradiation of acoustic waves, e.g. sonar systems
    • G01S15/88—Sonar systems specially adapted for specific applications
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/18—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using ultrasonic, sonic, or infrasonic waves

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Acoustics & Sound (AREA)
  • Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)

Abstract

The invention discloses a low-power-consumption underwater acoustic target detection system suitable for a small-sized platform. The invention is suitable for small-sized platforms, has the characteristics of low power consumption and small size, and can realize long-term large-scale marine observation; the target information can be provided for various underwater unmanned aircrafts and the like, and a foundation is laid for networking cooperative detection of the unmanned underwater vehicles; meanwhile, the target detection function of narrow-band and broadband sonar is realized, and the type of the target can be identified while the existence and the direction of the target are detected; the method can work in a non-stationary time-varying noise environment, and the detection threshold is calculated in a self-adaptive mode. The hardware of the invention is realized by adopting a CPU and an FPGA of a multi-chip packaging technology; the adopted CPU integrates large-capacity storage, thereby reducing the complexity of a circuit system and reducing the volume of an electronic system; the system is integrally realized by the low-power-consumption CPU and the FPGA, and compared with a detection system realized by adopting a traditional DSP chip, the overall power consumption is greatly reduced.

Description

Low-power-consumption underwater acoustic target detection system suitable for small platform
Technical Field
The invention belongs to the field of underwater target detection, and particularly relates to a low-power-consumption underwater acoustic target detection system suitable for a small-sized platform.
Background
Underwater target detection has wide application in the fields of civil use, national defense safety and the like. Because electromagnetic waves are greatly attenuated in water, underwater targets are generally detected by using an acoustic means. The traditional acoustic target detection equipment is large in size, high in energy consumption and low in intelligent degree. Generally, a conventional target detection system is mounted on a large-scale platform such as a ship or a warship, and first, acoustic signals are collected through a large-aperture array, and then, monitoring and identification are manually performed through an operator. An operator firstly judges whether a target appears according to the received signal, then determines the azimuth angle of the target, and finally identifies the target according to the signal characteristics.
With the development trend of miniaturization, intellectualization and networking of unmanned platforms, the application of target detection in various small-sized underwater/water surface unmanned platforms draws more and more attention. The small platform can be observed in the ocean for a long time because the small platform does not need to be operated by personnel.
The traditional detection system is directly applied to a small platform, and the defects of low autonomy, high energy consumption and the like of the detection method exist. The low autonomy reduces the detection and target parameter estimation performance of the small platform, and the long-term endurance of the small platform is reduced due to the excessive energy consumption. Therefore, an underwater acoustic array target detection system which is low in power consumption, high in autonomy and suitable for a small platform is extremely necessary.
Disclosure of Invention
The invention aims to provide a low-power-consumption underwater acoustic target detection system suitable for a small-sized platform aiming at the defects of the prior art. The invention realizes the detection, orientation and identification of the target on the target detection algorithm; the embedded real-time processing and low-power consumption long-term watching miniaturization and low-power consumption hardware circuit is realized on hardware.
The purpose of the invention is realized by the following technical scheme: a low-power-consumption underwater acoustic target detection system suitable for a small-sized platform comprises a preposed conditioning filtering module, an A/D conversion module, a CPU and an FPGA; the CPU comprises a single-channel energy testing module, an FFT module, a threshold detection module and a cluster identification module; the FPGA comprises a beam forming module and a median filtering module;
the preposed conditioning filtering module: filtering low-frequency and high-frequency interference signals in the analog signals of the N channels;
an A/D conversion module: converting the analog signals of the N channels output by the prepositive conditioning and filtering module into digital signals;
single channel energy detection module: the single-channel energy detection module extracts a channel from the digital signal output by the A/D conversion module and detects the energy change of the signal;
an FFT module: when the single-channel energy detection module detects that the signal energy of a certain channel exceeds a threshold value, the FFT module performs Fourier transform on data of all channels and outputs frequency domain data;
a beam forming module: performing frequency domain beam forming on the frequency domain data output by the FFT module to obtain an angle-frequency plane;
a median filtering module: estimating the background noise of the angle-frequency plane output by the beam forming module;
a threshold detection module: calculating a detection threshold plane at the current moment according to the estimated background noise output by the median filtering module, and searching for a suspicious value to obtain a detection result point set;
a clustering identification module: and distinguishing the target and outputting a target identification result comprising the target direction, frequency and target type according to the detection result point set output by the threshold detection module.
Further, the CPU also includes LPDDR2 and a multimedia instruction set; when the FFT module is used for operation, the multimedia instruction set is used for accelerating the FFT operation speed; LPDDR2 provides CPU with memory for calculation; the FPGA also comprises an RAM, and the RAM provides a calculation memory for the FPGA; the A/D conversion module consists of a plurality of low-power-consumption and low-noise A/D chips; the digital signal of N passways that A/D conversion module outputs inserts CPU in the form of the bus.
Furthermore, in the single-channel energy detection module, the extracted channel firstly passes through a band-pass filter bank and then carries out energy detection.
Further, when the single-channel energy detection module does not detect the target, the FFT module, the threshold detection module, the cluster identification module, the beam forming module and the median filtering module are closed.
Further, in the FFT module, inkThe first of time of day receptionnSignals of the channels n,k (t) The vector signal after Fourier transform isG n,k (f) The following formula:
Figure 243725DEST_PATH_IMAGE001
wherein the content of the first and second substances,n=1~N，findicating the scanning frequency.
Furthermore, in the beam forming module, the frequency domain signals of N channels are processedG n,k (f) Performing azimuth beam scanning to obtain an angle-frequency planeB k (θ,f) At the same time willB k (θ,f) Store in buffer queue
Figure 895286DEST_PATH_IMAGE002
(ii) a The formula is as follows:
B k (θ,f)=w(θ,f)H·G k (f)
wherein the content of the first and second substances,θwhich represents the angle of the scan, is,w(θ,f) Is the scan vector, and the superscript H denotes the conjugate transpose.
Further, in the median filtering module, estimation is carried out through vector signal processingkThe background noise of the angle-frequency plane at the moment is calculated by the following steps:
Figure 347127DEST_PATH_IMAGE003
wherein the content of the first and second substances,
Figure 227490DEST_PATH_IMAGE004
for the estimated background noise level, mid represents median filtering,
Figure 713966DEST_PATH_IMAGE005
as the queue length, | ·| non-conducting phosphor2Representing a 2-norm.
Further, the threshold detection module comprises:
(1) based on estimated background noise
Figure 723379DEST_PATH_IMAGE004
ComputingkTime of day detectionThreshold planeT k (θ,f)：
Figure 724833DEST_PATH_IMAGE006
Wherein the content of the first and second substances,χ 2the distribution of the Chinese character 'chi' square is,P fa is the expected false alarm probability;
(2) traverse angle-frequency planeB k (θ,f) Find out to satisfyB k (θ,f)>T k (θ,f) And forming a set;Ia set of detection result points consisting of peak points as
Figure 392575DEST_PATH_IMAGE007
，i=1,…,IWherein
Figure 481360DEST_PATH_IMAGE008
And
Figure 209145DEST_PATH_IMAGE009
respectively representiAzimuth and frequency for each peak.
Further, the threshold detection module calculates in the traversal angle-frequency planeT k (θ,f) While simultaneously comparingB k (θ,f)>T k (θ,f) And the detection result can be obtained only by traversing once.
Further, the cluster identification module comprises:
(a) for the detection result point set
Figure 635578DEST_PATH_IMAGE007
Carrying out mean shift clustering on the internal signals according to angles to obtainLClusters, each cluster representing one target, each cluster f l Consisting of close angles and corresponding frequencies,l=1~L；
(b) each cluster r l The frequency of the cluster is matched and identified with the features in the target feature library Ψ to obtain the target type represented by each cluster, and the specific identification method is as follows:
Figure 90699DEST_PATH_IMAGE010
wherein the content of the first and second substances,
Figure 82926DEST_PATH_IMAGE011
for saving the firstlThe target type of the cluster to be identified finally;
Figure 247191DEST_PATH_IMAGE012
for the number of frequencies in each cluster,
Figure 957658DEST_PATH_IMAGE013
is as followslIn a clusterpThe frequency of each of the plurality of frequencies,αto scale factor, Ψ m,j Is the first in ΨmThe second of the candidate object typejA characteristic frequency.
The invention has the beneficial effects that:
(1) the invention is suitable for small-sized platforms, has the characteristics of low power consumption and small size, and can realize long-term large-scale marine observation;
(2) the invention can provide target information for various underwater unmanned aircrafts and the like, and lays a foundation for networking cooperative detection of unmanned underwater vehicles;
(3) the invention simultaneously realizes the target detection function of narrow-band and broadband sonar, and can identify the type of the target while detecting the existence and the direction of the target;
(4) the invention can work in the non-stationary time-varying noise environment and adaptively calculate the detection threshold;
(5) the hardware of the invention is realized by adopting a CPU and an FPGA of a multi-chip packaging technology, the CPU realizes FFT acceleration by using a self-contained multimedia instruction set in the operation process, and data buffering is realized by the LPDDR2 integrated with the CPU in the operation process, thereby improving the calculation efficiency;
(6) the CPU adopted by the invention integrates large-capacity storage, thereby reducing the complexity of a circuit system and reducing the volume of an electronic system;
(7) the data buffering in the beam forming and median filtering processes is realized by an RAM inside an FPGA, the whole system is realized by a low-power-consumption CPU and the FPGA, and compared with a detection system realized by adopting a traditional DSP chip, the whole power consumption is greatly reduced.
Drawings
One or more embodiments are illustrated by way of example in the accompanying drawings, which correspond to the figures in which like reference numerals refer to similar elements and which are not to scale unless otherwise specified.
FIG. 1 is a block diagram of a low power consumption underwater acoustic target detection system of the present invention;
FIG. 2 is a diagram illustrating a signal detection result at a certain time according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a target recognition result according to an embodiment of the present invention.
Detailed Description
In order to meet the requirement of low power consumption, the hardware part of the invention is realized by adopting a low-power consumption Central Processing Unit (CPU) based on a multi-chip packaging technology and a low-power consumption Field Programmable Gate Array (FPGA) based on a flash architecture; the CPU and the FPGA are connected by adopting a parallel bus.
As shown in FIG. 1, the invention relates to a low-power-consumption underwater acoustic target detection system suitable for a small platform, which comprises a preposed conditioning filtering module, an A/D conversion module, a CPU and an FPGA. The CPU comprises a single-channel energy testing module, an FFT module, a multimedia instruction set, a threshold detection module, a cluster identification module and a low voltage memory rate 2 (LPDDR 2) memory; the FPGA includes a beamforming module and a median filtering module and Random Access Memory (RAM). LPDDR2 provides the CPU with computational memory and RAM provides the FPGA with computational memory.
(1) Assuming that the acoustic array carried on the small platform has N channels, firstly, acoustic analog signals of the N channels of the array pass through a preposed conditioning filtering module to filter low-frequency and high-frequency interference signals.
(2) And converting the analog signals of the N channels output by the preposed conditioning filtering module into digital signals through an A/D conversion module. The A/D conversion module is composed of a plurality of low-power-consumption and low-noise A/D chips.
(3) The digital signal of N passways that A/D conversion module outputs inserts CPU in the form of the bus. In order to reduce the calculation power consumption, when a target is not detected, a signal of one channel is extracted by a CPU through a single-channel energy detection module to carry out real-time detection, and the energy change of the signal is detected by adopting a conventional energy detection method. The extracted channel firstly passes through a band-pass filter bank and then single-channel energy detection is carried out. The filter bank can improve the detection performance of weak signals.
When the single-channel energy detection module detects that the signal energy of a certain channel exceeds a threshold value, the FFT module performs Fourier transform (FFT) on data of all channels, otherwise all subsequent detection algorithm modules are closed, wherein the subsequent detection algorithm modules comprise an FFT module, a threshold detection module, a cluster identification module, a beam forming module and a median filtering module, the single-channel energy detection module is only used for guarding, and the power consumption of the system is reduced.
When FFT operation is carried out, the multimedia instruction set of the selected CPU is utilized to accelerate the FFT operation speed. Under some special application scenes (such as high-risk target searching, platform energy abundance and the like), all subsequent detection algorithm modules can be continuously started, and the overall target detection and identification efficiency of the system is improved under the condition of sacrificing energy consumption.
In thatkThe first of time of day receptionnSignals of the channels n,k (t) The vector signal after time-frequency conversion isG n,k (f) The following formula:
Figure 655618DEST_PATH_IMAGE001
wherein the content of the first and second substances,n=1~N，findicating the scanning frequency.
(4) And the FFT module outputs frequency domain data to a beam forming module of the FPGA to complete frequency domain beam forming.
For frequency domain signals of N channelsG n,k (f) Performing azimuth beam scanning to obtain an angle-frequency planeB k (θ,f) At the same time willB k (θ,f) Store in buffer queue
Figure 767930DEST_PATH_IMAGE002
(ii) a The concrete formula is as follows:
B k (θ,f)=w(θ,f)H·G k (f)
wherein the content of the first and second substances,θwhich represents the angle of the scan, is,w(θ,f) Is the scan vector, and the superscript H denotes the conjugate transpose. The signal receiving signal-to-noise ratio can be improved by the beam forming, and the performance gain of subsequent detection and identification is improved.
(5) And estimating the background noise of the angle-frequency plane output by the beam forming module by using a median filtering module, and outputting the result to the CPU.
Estimation by vector signal processingkThe background noise of the angle-frequency plane at the moment is calculated by the following steps:
Figure 837517DEST_PATH_IMAGE003
wherein the content of the first and second substances,
Figure 284548DEST_PATH_IMAGE004
for the estimated background noise level, mid represents median filtering,
Figure 832204DEST_PATH_IMAGE005
as the queue length, | ·| non-conducting phosphor2Representing a 2-norm.
(6) The CPU threshold detection module calculates the current detection threshold plane according to the estimated background noise output by the median filtering module in one traversalT k (θ,f) And to look for a suspect value,and the calculation amount and the memory consumption are saved.
(6.1) estimating the background noise
Figure 2285DEST_PATH_IMAGE004
Computing kTime of day detection threshold planeT k (θ,f)：
Figure 990577DEST_PATH_IMAGE006
Wherein the content of the first and second substances,χ 2the distribution of the Chinese character 'chi' square is,P fa is the expected false alarm probability.
(6.2) traversing the angle-frequency planeB k (θ,f) Find out to satisfyB k (θ,f)>T k (θ,f) And form a set. Assume consensus IA peak value of IA set of detection result points consisting of peak points as
Figure 675636DEST_PATH_IMAGE007
，i=1,…,IWherein
Figure 26983DEST_PATH_IMAGE008
And
Figure 300838DEST_PATH_IMAGE009
respectively representiAzimuth and frequency for each peak. Due to the influence of environmental noise and mechanical noise of the underwater platform, the set may contain detection results caused by random interference.
In practical implementation, the CPU calculates in traversing angle-frequency planeT k (θ,f) At the same time compareB k (θ,f)>T k (θ,f) The detection result can be obtained only by traversing once without traversing calculation in advance and storingT k (θ,f) Meanwhile, the calculation amount of the algorithm and the memory consumption are saved.
(7) And the cluster identification module of the CPU distinguishes the target and outputs a target identification result according to the detection result point set output by the threshold detection module.
(7.1) set of detection result points
Figure 712228DEST_PATH_IMAGE007
The mean shift clustering is carried out on the signals in the space according to the angles, and the signals can be obtained after the clustering is setLClusters, each cluster representing one target, each cluster f l Consisting of close angles and corresponding frequencies,l=1~L。
(7.2) dividing each cluster r l The frequency of the cluster is matched and identified with the features in the target feature library Ψ to obtain the target type represented by each cluster, and the specific identification method is as follows:
Figure 884583DEST_PATH_IMAGE014
wherein the content of the first and second substances,
Figure 774042DEST_PATH_IMAGE011
for saving the firstlThe target type that is ultimately identified for each cluster.
Figure 669448DEST_PATH_IMAGE012
For the number of frequencies in each cluster,
Figure 517318DEST_PATH_IMAGE013
is as followslIn a clusterpThe frequency of each of the plurality of frequencies,αto scale factor, Ψ m,j Is the first in ΨmThe second of the candidate object typejA characteristic frequency;
Figure 363920DEST_PATH_IMAGE015
is associated with the target type, provided by the target feature library Ψ.
(7.3) outputting detection and identification results including the target azimuth
Figure 57070DEST_PATH_IMAGE008
Frequency of
Figure 852987DEST_PATH_IMAGE009
And type of object
Figure 606180DEST_PATH_IMAGE011
。
Taking the system of the present invention mounted on an AUV as an example, the AUV performs a target search task in a certain water area. There are three targets in the water domain, 2 targets of type a and 1 target of type B. Different types of targets have a range of different line spectral features.
At some point, the actual azimuth angles for type A targets are about-60 and-23, and type B targets are located about 29.
At this time, the signals collected by the system pass through an angle-frequency plane obtained in the FPGAB k (θ,f) As shown in fig. 2, where the x-axis is the scan angle and the y-axis is the frequency.
After the CPU runs the detection algorithm, the system detects the signal point set
Figure 462051DEST_PATH_IMAGE007
Marked with an "x" in figure 2. It can be seen that a series of signals with a certain regularity of frequency are detected in the vicinity of-60 °, -23 °, 29 °, while three randomly distributed noise interferences are present.
The detection result is compared with the target feature library Ψ by a cluster identification module in the CPU, and the obtained target number and identification result are shown in fig. 3. For convenience of illustration, fig. 3 shows different target types with different marks, and the recognition results are also marked near the respective detection points. The system identifies 3 targets in total, 2 targets of type a and 1 target of type B, and other random interferences are also marked separately. It can be seen that the results of the target type, the target orientation, etc. are all consistent with the actual situation.
In the embodiment, because the low-power-consumption central processing unit and the FPGA which adopt the multi-chip packaging technology are adopted, the detection system is small in size and can be directly carried on various types of small Autonomous Underwater Vehicles (AUV), underwater gliders and the like, such as a portable small AUV of 50 kg grade and a swallow underwater glider of 70 kg grade.
In the aspect of using power consumption, the self endurance time of a certain domestic 50 kg-level AUV is about 5 hours, and the endurance time of a system carrying the system is about 5 hours. The swallow glider continues the voyage for about 90 days, and the subsequent voyage time of carrying the invention is about 90 days, so that the system has extremely low power consumption and almost no influence.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions.

Claims (9)

1. A low-power-consumption underwater acoustic target detection system suitable for a small-sized platform is characterized by comprising a preposed conditioning filtering module, an A/D conversion module, a CPU and an FPGA; the CPU comprises a single-channel energy testing module, an FFT module, a threshold detection module and a cluster identification module; the FPGA comprises a beam forming module and a median filtering module;
the preposed conditioning filtering module: filtering low-frequency and high-frequency interference signals in the analog signals of the N channels;
an A/D conversion module: converting the analog signals of the N channels output by the prepositive conditioning and filtering module into digital signals;
single channel energy detection module: the single-channel energy detection module extracts a channel from the digital signal output by the A/D conversion module and detects the energy change of the signal; when the single-channel energy detection module does not detect the target, the FFT module, the threshold detection module, the cluster identification module, the beam forming module and the median filtering module are closed;
an FFT module: when the single-channel energy detection module detects that the signal energy of a certain channel exceeds a threshold value, the FFT module performs Fourier transform on data of all channels and outputs frequency domain data;
a beam forming module: performing frequency domain beam forming on the frequency domain data output by the FFT module to obtain an angle-frequency plane;
a median filtering module: estimating the background noise of the angle-frequency plane output by the beam forming module;
a threshold detection module: calculating a detection threshold plane at the current moment according to the estimated background noise output by the median filtering module, and searching for a suspicious value to obtain a detection result point set;
a clustering identification module: and distinguishing the target and outputting a target identification result comprising the target direction, frequency and target type according to the detection result point set output by the threshold detection module.
2. A low power consumption underwater acoustic target detection system suitable for use with small platforms as claimed in claim 1 wherein the CPU further includes LPDDR2 and a multimedia instruction set; when the FFT module is used for operation, the multimedia instruction set is used for accelerating the FFT operation speed; LPDDR2 provides CPU with memory for calculation; the FPGA also comprises an RAM, and the RAM provides a calculation memory for the FPGA; the A/D conversion module consists of a plurality of low-power-consumption and low-noise A/D chips; the digital signal of N passways that A/D conversion module outputs inserts CPU in the form of the bus.
3. A low power consumption underwater acoustic target detection system suitable for small platforms as in claim 1 wherein in the single channel energy detection module, the extracted channel passes through a band pass filter bank before energy detection.
4. A low power consumption underwater acoustic target detection system adapted for use with a compact platform as claimed in claim 1 wherein in the FFT module, inkThe first of time of day receptionnSignals of the channels n,k (t) The vector signal after Fourier transform isG n,k (f) The following formula:
Figure 993969DEST_PATH_IMAGE002
wherein the content of the first and second substances,n=1~N，findicating the scanning frequency.
5. The system of claim 4, wherein the beamforming module is configured to perform beamforming on the N channels of frequency domain signalsG n,k (f) Performing azimuth beam scanning to obtain an angle-frequency planeB k (θ,f) At the same time willB k (θ,f) Store in buffer queue
Figure 707847DEST_PATH_IMAGE004
(ii) a The formula is as follows:
B k (θ,f)=w(θ,f)H·G k (f)
wherein the content of the first and second substances,θwhich represents the angle of the scan, is,w(θ,f) Is the scan vector, and the superscript H denotes the conjugate transpose.
6. A low power consumption underwater acoustic target detection system suitable for small platforms as in claim 5 wherein the median filtering module estimates by vector signal processingkThe background noise of the angle-frequency plane at the moment is calculated by the following steps:
Figure 18743DEST_PATH_IMAGE006
wherein the content of the first and second substances,
Figure 210690DEST_PATH_IMAGE008
for the estimated background noise level, mid represents median filtering,
Figure 556220DEST_PATH_IMAGE010
as the queue length, | ·| non-conducting phosphor2Representing a 2-norm.
7. A low power consumption underwater acoustic target detection system adapted for use with small platforms as recited in claim 6 in which the threshold detection module comprises:
(1) based on estimated background noise
Figure 175421DEST_PATH_IMAGE008
ComputingkTime of day detection threshold planeT k (θ,f)：
Figure DEST_PATH_IMAGE012
Wherein the content of the first and second substances,χ 2the distribution of the Chinese character 'chi' square is,P fa is the expected false alarm probability;
(2) traverse angle-frequency planeB k (θ,f) Find out to satisfyB k (θ,f)>T k (θ,f) And forming a set;Ia set of detection result points consisting of peak points as
Figure DEST_PATH_IMAGE014
，i=1,…,IWherein
Figure DEST_PATH_IMAGE016
And
Figure DEST_PATH_IMAGE018
respectively representiAzimuth and frequency for each peak.
8. A low power consumption underwater acoustic target detection system suitable for small platforms as in claim 7 wherein the threshold detection module calculates in the traversal angle-frequency planeT k (θ,f) While simultaneously comparingB k (θ,f)>T k (θ,f) And the detection result can be obtained only by traversing once.
9. The system of claim 7, wherein the cluster identification module comprises:
(a) for the detection result point set
Figure 157633DEST_PATH_IMAGE014
Carrying out mean shift clustering on the internal signals according to angles to obtainLClusters, each cluster representing one target, each cluster f l Consisting of close angles and corresponding frequencies,l=1~L；
(b) each cluster r l The frequency of the cluster is matched and identified with the features in the target feature library Ψ to obtain the target type represented by each cluster, and the specific identification method is as follows:
Figure DEST_PATH_IMAGE020
wherein the content of the first and second substances,
Figure DEST_PATH_IMAGE022
for saving the firstlThe target type of the cluster to be identified finally;
Figure DEST_PATH_IMAGE024
for the number of frequencies in each cluster,
Figure DEST_PATH_IMAGE026
is as followslIn a clusterpThe frequency of each of the plurality of frequencies,αto scale factor, Ψ m,j Is the first in ΨmThe second of the candidate object typejA characteristic frequency.
CN202111348394.6A 2021-11-15 2021-11-15 Low-power-consumption underwater acoustic target detection system suitable for small platform Active CN113777610B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111348394.6A CN113777610B (en) 2021-11-15 2021-11-15 Low-power-consumption underwater acoustic target detection system suitable for small platform

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111348394.6A CN113777610B (en) 2021-11-15 2021-11-15 Low-power-consumption underwater acoustic target detection system suitable for small platform

Publications (2)

Publication Number Publication Date
CN113777610A CN113777610A (en) 2021-12-10
CN113777610B true CN113777610B (en) 2022-04-01

Family

ID=78873894

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111348394.6A Active CN113777610B (en) 2021-11-15 2021-11-15 Low-power-consumption underwater acoustic target detection system suitable for small platform

Country Status (1)

Country Link
CN (1) CN113777610B (en)

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB0802936D0 (en) * 2008-02-18 2008-06-04 Curtis Thomas E Underwater Surveillance
US9869752B1 (en) * 2016-04-25 2018-01-16 Ocean Acoustical Services And Instrumentation Systems, Inc. System and method for autonomous joint detection-classification and tracking of acoustic signals of interest
CN109856593B (en) * 2018-12-21 2023-01-03 南京理工大学 Sound source direction-finding-oriented miniature intelligent array type acoustic sensor and direction-finding method thereof
CN113126099B (en) * 2020-01-14 2023-05-02 中国科学院声学研究所 Single-beam substrate classification method based on audio feature extraction
CN111580077B (en) * 2020-05-29 2022-09-16 中国电子科技集团公司第三研究所 Early warning equipment and method for detecting small target sonar

Also Published As

Publication number Publication date
CN113777610A (en) 2021-12-10

Similar Documents

Publication Publication Date Title
US10725149B1 (en) System and method for autonomous joint detection-classification and tracking of acoustic signals of interest
Duan et al. Comments on" unbiased converted measurements for tracking"
CN111024209B (en) Line spectrum detection method suitable for vector hydrophone
CN110007299B (en) Weak target detection tracking method based on mixed coordinate pseudo-spectrum technology
CN105137423A (en) Real-time detection and separation method of multiple moving objects by through-the-wall radar
CN107632291A (en) A kind of method based on the anti-corner reflector interference of polarimetric radar
CN109521430A (en) A kind of range extension target detection method inhibiting narrowband interference
CN110109068B (en) Foil strip interference identification method based on polarization characteristic vector
CN113777610B (en) Low-power-consumption underwater acoustic target detection system suitable for small platform
CN109829416B (en) Method and system for detecting ultrahigh frequency partial discharge signal from random noise
Zhang et al. Underwater target detection based on machine learning
CN113009443B (en) Sea surface target detection method and device based on graph communication density
CN114690143B (en) Method and device for suppressing radar clutter, radar and medium
CN115508791A (en) Intelligent target fusion detection method in unknown dry noise environment
CN116299208A (en) Anti-interference method based on active/passive radar composite guide head data association
CN103424738A (en) Reverberation suppression method based on morphology
Yang et al. Classification of ship-radiated signals via chaotic features
CN203149113U (en) Substation abnormal noise monitoring and positioning instrument
Zhan et al. Research on KL transform based on wavelet transform for anti-chaff-jamming
Faramarzi et al. Infrared dim and small targets detection based on multi-scale local contrast measure utilizing efficient spatial filters
Liang et al. Detection of narrow-band sonar signal on a Riemannian manifold
CN103487794A (en) Underwater reverberation suppression method based on wavelet packet transforming
CN117930208A (en) Anti-platform interference pulse signal detection line spectrum extraction method for far-field target airspace characteristics
CN113341393B (en) Method for detecting weak target of radar without direct wave channel external radiation source
CN114114222B (en) Broadband target detection method under strong interference complex environment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant